86 ALASKAN GLACIER STUDIES
the mountains on which, for some reason, little d6bris exists. This is typically illustrated in the Variegated Glacier in this region and in the Allen Glacier in the Copper Biver region. Such a clear ice area developed in Atrevida Glacier in 1906, as already stated. We infer the existence of a similar clear ice area in the Galiano Glacier, which by more rapid ablation was reduced to the condition first of an interior flat and then, when low enough, to a region of alluviation. It was in the latter condition in 1890, and was then receiving deposits not only from Galiano Stream, but also from the drainage of Atrevida Glacier.
The large ice block which we believe to be buried here must in the course of time receive a thick cover of outwash gravels which have a tendency to preserve it from wastage. But inevitably, in the course of time, if stagnation continues, this ice will completely melt away. If alluviation continues no depression may result, but if glacial streams cease to bring sediment, a large kettle would ultimately mark the site of the buried ice or, since the ocean is so near, an arm of the sea. The present condition and possible future of this region have more than local interest. It is one of many instances of more or less completely-buried masses of stagnant ice in the Alaskan region, which show us in actual existence what was undoubtedly common along the margin of the waning continental ice masses of the Glacial Period. The literature of glacial phenomena does not assign to buried masses of stagnant ice the full importance which they probably had in shaping the topography of marginal deposits. The conditions near Galiano Glacier, and elsewhere in the Yakutat Bay region, give facts which ought to be of value in interpretation of these phenomena.
Condition in 1909. The interval between the visits of 1905 and 1906 was too brief for any notable change in the condition of Galiano Glacier and the region to the south of it. But by 1909 there had been so much change that it was noticeable, and this change was of several kinds. The area of the ablation moraine in the mountain valley had extended farther toward the valley head than in 1905, when it evidently did not extend so far as in 1890. Alder had again spread over a large part of the piedmont area (PL XXXII, A) and even into the mountain valley. In most places the alder was young and in scattered individuals and clusters; but in the more stable parts of the piedmont area, and especially on the eastern side, there were already alder thickets through which it was difficult to travel, and in which the bushes were eight or nine years old. There was still much slumping of moraine in the piedmont area, probably because crevasses still existed there, and there were large pits barren of vegetation, but surrounded by alder growth; and there were other places where recent subsidence had overturned the alder that had begun to grow. Evidently the glacier had not yet settled down to a state of stability equal to that which permitted the growth of such a uniform cover of alder as was present in 1890.
There were also notable changes in the marginal valleys on both the east and west sides. That on the east side was much broader than in 1905 and headed farther up in the mountain valley. Only a small stream emerged from the glacier on this side, as was true also in 1905; but it was cutting the glacier back, forming a steep ice cliff which was extending down-stream, and thus was broadening its valley. At its lower end the ice cliff merged into a moraine-covered ice embankment on which young alders were growing. At this point, where the ice cliff was being extended down-stream, the moraine and its alder growth were sliding down the slope and here one saw withered, present-